It’s been an exceptionally busy quarter for the AMC Health analytics team. From groundbreaking dashboards to pioneering machine learning models, the team is resetting expectations – both internally and externally – on the value that robust data and analytics bring to patient health.
Transforming Business Intelligence to Better Serve
We have a transformational Service Operations KPI dashboard that offers unprecedented insights into our own operations, showcasing data in ways we've never seen before. It uncovers detailed patterns in service ticket lifecycles and device swap-out severity, drawing on previously inaccessible sources. This dashboard is not just about tracking metrics; it’s designed to enhance our training for new operations staff and boost efficiency across departments. As we scale our Commercial accounts, this tool is a critical asset, providing essential support to the clinical team in our quest to become a world-class service provider.
Advancements in Machine Learning Focused on Results
Our data science team has just completed an advanced model to predict which veterans are at risk of premature disenrollment—those on long-term monitoring programs who might leave our program within the first 90 days. This is critical to improving clinical outcomes, as our Commercial data clearly shows that long-term, chronic care patients achieve better results when they stay on programs longer.
Our final results surpassed expectations, showing the highest predictive accuracy of any model we've developed thus far. The next steps involve creating automated reports to identify at-risk veterans, collaborating with the clinical team to refine training for VA support staff and Care Coordinators, and targeting specific VISNs where early churn is a major issue.
Harnessing Conversational Data for Enhanced Enrollment Success
On another front, we’re diving into machine learning models to enhance our Commercial candidate enrollment strategies. We’ve started analyzing vast amounts of unstructured data from our call center platform. This analysis is revealing crucial insights, such as the optimal talking speed for onboarding patients, key phrases to avoid, and the critical window—just 60 seconds—that our agents have to persuade candidates.
An exciting development is our upcoming experiment with Large Language Models (LLMs), like those behind ChatGPT. We’re using these advanced tools to analyze conversational data and fine-tune our messaging strategies. By identifying the most effective topics, expressions, and arguments, we aim to significantly improve our candidate engagement and conversion rates.
The AMC Health analytics team is leading the way in innovation. We’re excited about these developments and look forward to sharing more as we continue to push the envelope. Stay tuned for more updates on how we’re leveraging data to drive success and make a real impact.